Hiho,

when using the Z scripts they 'learn' when running and taking trades. When stopping the strategy this knowledge will be lost cause it seems only to be held in memory.

@JCL: You said that this will be changed in future versions. Can we hope to see this in V1.2 ?

No, 1.2 is finished already, but it is on our list for a future version.

Oh. Mmmhh. Okay. Can you tell me how big the impact would be when a running Z12 version (2 month) will be continued with v1.2 in performance? I can't estimate if it makes sense to stay on V1.16 for that impact...

jcl,

Am I correct in saying that this only has an impact if phantom is enabled, as otherwise the strategies ignore the equity curve of the algos? and that z12 is reoptimised with new parameters and Factors before release?

Crucially the Factors should reflect the gist of what the phantom trading/equity curves is telling us about the profitability of the asset algo combos over the period?

Interesting that we are discussing this topic, because I was just about to post an idea related to this...

For a future feature request... can you make it possible for Zorro to run a simulation for a (specified) period of the start of Trade, so that it can already align itself for the current live trading market, with regard to phantom trades and any other learned stats?

A similar feature which you already have planned, I believe, is the one related to filling the Lookback with (some) history data at the time of pressing Trade, rather than 100% live from the broker feed.

So this idea would be to first ensure there is no gap between the end of history data, and the live available feed data. Then run simulation for (specified) period, causing Zorro to establish some trades and equity curve. Then seamlessly continue from the live trading feed.

That all sounds straightforward when I say it out loud, but I'm sure its a technical feat!!
Thanks